We are proposing a cross-sectional, longitudinal, field-based study of the development of children's self and task perceptions, self-esteem, and activity preferences across three activity domains common to childhood experience: academic, social and physical domains. Our previous work in the academic achievement domain has indicated that both parents' beliefs and children's perceptions of their competence expectations for success, task difficulty perceptions, and task value perceptions are critical mediators of achievement behavior and choices among children in grades 5-12. We propose to extend this work to younger children and a broader set of children's activities to study four basic issues: (1) the development of self and task beliefs within and across domains, (2) the role of these beliefs in shaping children's behavioral choices across the domains, (3)\the antecedents of parents' and teachers' beliefs about their children in each of these domains, and (4) the impact of parenting and teaching styles and of teacher and parent beliefs, values, and perceptions on children's developing self and task beliefs. Subjects will be tested five times over a four year period. Initially, approximately 1000 students in grades K-2 and their families will be recruited. Teacher ratings of student competence in each domain and children's performance on standardized measures of mental and physical abilities will be obtained at Wave 0 (Spring 1986). Parents' perceptions and beliefs will also be obtained at Wave 0. At waves 1 (Spring, 1987), 2 (Fall, 1987), 3 (Spring, 1988), and 4 (Spring 1989), both questionnaire and interview procedures will be used to measure the self and task beliefs of the children, and their teachers. Teacher estimates of competence and standardized competency tests will also be given. In addition, sociometric measures will be used to get peer perceptions of each child's social competence. Parents' beliefs and perceptions will be assessed in waves 1, 3, and 4. Two of the waves (2 and 3) will be gathered at the beginning and end of a school year in order to model within year teacher/classroom effects. Both structural equation modelling procedures and more traditional statistical procedures will be used to assess the developmental changes in children's, parents', and teachers' beliefs and the relations among these changes.